72

Lymphology 16 (1983) 72-82 ©Georg Thieme Verlag Stuttgart · New York

lmmunoarchitecture of the Human

Th.M. Grogan, M.D., C.S. Jolley, M.T. (ASCP), C.S. Rangel, H.T.L. (ASCP)

Department of Pathology, University of Arizona, Arizona Health Sciences Center, Tucson, Arizona

Summary plastic involvement of the spleen was absent New immunochemical staining techniques reveal by both gross and microscopic inspection in­ key topographic aspects of splenic immune function. dicating normal immunoreactive , al­ Distinct compartmentalization of B and T cells and lymphocytic subtypes within the white and red though we cannot exclude minor pathologic pulp of the human spleen provide a morphologic alterations. Resected specimens were snap basis for division of labor in the immune response. frozen in OCT compound (Ames Company Division of Miles Laboratory, Elkart, Indiana) and stored at - 70 °C. Introduction Tissue section analysis was performed using The human spleen functions as an immuno­ biotin-avidin conjugation and horseradish per­ logic ftlter of the bloodstream. Although oxidase labelling with diarninobenzidine tetra­ its importance in immune function has hydrochloride as the detection agent (1, 3). long been recognized and measured, details Briefly, cryostat sections of splenic tissues of splenic immunoreactivity remain elusive were fixed in acetone at 4 °C for 10 minutes, in part because topographic analysis of air-dried at room temperature, and then hy­ splenic immunologic features has heretofore drated in PBS at pH 7.4 for 10 minutes. Fol­ not been possible. Previously, red and white lowing this wash, mouse monoclonal antibody pulp cells were admixed in cell suspension to human cell surface antigen was applied assays that disrupted architecture. The ad­ (Table). vent of tissue section immunologic assessment After a second PBS wash, a second stage rea­ with hybridoma-derived monoclonal antibod­ gent, biotin conjugated F(ab')2 goat anti­ ies, a biotin-avidin conjugating system and mouse IgG (Tago, Incorporated) was added peroxidase labelling, now makes possible topo­ followed by 15 minute incubation and a 3 graphic study of immunologic function and de­ minute PBS wash. The cryostat section was lineation of the immunoarchitecture of lym­ next incubated with avidin D conjugated with phoreticular organs (1, 2). The present study horseradish peroxidase (Vector Laboratories, applies these newer techniques and reagents Inc.). After another 15 minute incubation pe­ to examine B and T and histio­ riod, the section was further washed twice cytes within the splenic white and with PBS. Next, each section was incubated to determine if immunologic compartmentali­ with diaminobenzadine tetrahydrochloride zation occurs as in lymph nodes and the (DAB) solution consisting of three mg of DAB . per ml of PBS along with 0.010 ml of 30% H2 0 2 • After the 5 minute incubation in DAB Methods solution, sections were washed in PBS for 3 Six spleens were removed during staging lapa­ minutes and rinsed in distilled water. Slides rotomies for . In each instance neo- were incubated in a copper sulfate solution

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. lmmunoarchitecture of the Human Spleen 73

Table Hybridoma-derived monoclonal antibodies*

Monoclonal Hybridoma Specificity Abbreviation Concentration antibody clone

Leu 1 L17F12 Mature T-cells PAN-T 2 J.Lg/ml Leu 2a SK1 Suppressor-cytoxic Ts/c 2 J.Lg /ml T-cells Leu 3a SK3 Helper-inducer-T cells Th 2J.Lg /ml Leu 5 ATM 1.1 Sheep ABC ER 2 J.Lg /ml Receptor-bearing (ER) T-cells Leu 7 HNK-1 Natural killer cells NK 2 J.Lg /ml Anti-HLA-DR L243 H LA-DR antigen fl-lym- Ia 1 J.Lg /ml phocytes, monocytes/ histocytes, activated T-cells Anti-kappa 163-42 kappa-immunoglobulin lgK 0.5 J.Lg/ml light chains, cytoplasmic and surface Anti-lambda 1-155-2 lambda-immunoglobulin lg71. 0.5 J.Lg/ml light chains, cytoplas- mic and surface

*Obtained from Becton-Dickinson, Oxnard, Calif. for 5 minutes, rinsed with PBS, dipped in dis­ magnification, the preponderance of mature tilled water and counter-stained with methy­ T cells in the and the spillover into lene blue for 5 minutes, followed by a brief the red pulp. At higher magnification in Fig.2, 100% ethanol dip, two changes of xylene and T cells are identified by a darkly stained peri­ coverslipped with permount. phery in contrast to T antigen negative cells Alpha-naphthyl butyrate esterase activity indi­ without peripheral surface staining. A light cative of the presence of monocytic/histiocytic counterstain of methylene blue indicates in cells was detected on cryostat sections as previ­ gray tones the non-staining cellular compo­ ously described ( 4). nents seen predominantly in the germinal cen­ ter and in the red pulp. As indicated, an occa­ Results and Discussion sional extremely dark irregular cell with natural peroxidase activity is visible particularly in the As displayed in the figures, distinct antigenic red pulp. Figure 3 delineates graphically the staining patterns are clearly identifiable in the microanatomic features of this same cryostat splenic tissue sections using monoclonal anti­ section of splenic white pulp depicting the ana­ bodies with the biotin-avidin horseradish per­ tomic terms traditionally applied to the com­ oxidase technique. Our findings indicate that partments of the white pulp (2 , 5). Figs. 1- 3 compartmentalization of the B and T cells oc­ considered together illustrate the utility of the curs within the white pulp. Although B and tissue sections mode of immunologic assess­ areas are largely separate, some overlap ment and the sharp details of immunoarchitec­ of these zones exists with spillover of B and T ture discernible by this approach. cells into the red pulp. The staining pattern in Figures 1 and 2 demon­ strate a predominance of mature dark staining A Localization of mature T cells: T-cells in the periarteriolar lymphoid sheath Figures 1 and 2 demonstrate the distribution (PALS). This pattern signifies that mature T of mature T cells. Figure 1 emphasizes at low cells are largely compartmentalized in the PALS .

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. 74 Th.M. Grogan, C.S. Jolley, C.S. Rangel

Fig. 1 Localization of mature T cells using Leu 1 antibody. Darkly stained mature T cells predominate in the white pulp (WP) and spill over into the red pulp ( RP) . Magnification 25 x

They also appear in the adjacent of mature T cells in the marginal zone is of (MZ) and mantle of the . A interest because many splenic arterioles seem few are found scattered wjthin the germinal to terminate here (Figs. 2 and 3), and this is center and the red pulp. Thus, while substan­ also the site where antigen trapping takes place tial compartmentalization of mature, splenic (2, 5, 6). Some evidence suggests that antigen T cells may exist in the PALS and MZ zones, does not remain in the marginal zone but is there is intimate contact of T cells wjth B transferred to the germinal center where its cells in the germinal centers and with macro­ presence in dendritic reticulum cells stimulates phages in the red pulp. deveopment (5 - 7). It is possible that Mitoses identified in the PALS region of most some antigen is transported to the PALS re­ spleens examined indicate T cell proliferation gion to stimulate T cell clonal expansion. Al­ in this zone. Previously T cells in the PALS though this immunochemical technique cannot were thought to represent T cells migrating precisely delineate the sequence of events in from the bloodstream and "homing" to the antigen trapping and subsequent T & B cell spleen (2 , 5). The observed mitoses, however, clonal expansion, further application to in vivo suggest that T cell clonal expansion may also animal studies may pinpoint the sequential occur in this region substantiating local proli­ white pulp response (lymphocytopoiesis). feration as well as systemic migration. B.Localization ofT versus B cells: In all six spleens the T cells in the marginal zone do not display mitoses. Mature T cells Using the tissue section method of immunolo­ in this region presumably migrate from the gic assessment, sections can be examined for T PALS zones of clonal expansion or enter di­ & B cell antigens sequentially. The staining pat­ rectly from the circulation. The prominence tern in Fig. 4 demonstrates T & B compartmen-

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. lmmunoarchitecture of the Human Spleen 75

Fig. 2 Localization of mature T cells using Leu antibody. Darkly stained T cells predominate in the peri­ arteriolar regions and are scarce in the germinal center (GC) . The arrows indicate dark irregular cells with natural peroxidase activity. Magnification 250 x

;" ,.--- ...... ______... RP ./ ; .... ,.- .... ' ..... ______.,/ " MZ

PALS

;'------, ..... RP 'I / ..... \ / ', \ / ...... ' .... RP ---- ...... _.,,

Fig. 3 Microanatomic compartments of the white pulp. The anatomic features of the white pulp in Fig. 2 are delineated. PALS= periarteriolar lymphoid sheath; CA =central arteriole; GC =germinal center; MZ = marginal zone; M = mantle; A = arteriole; RP = red pulp

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. 76 Th.M. Grogan, C.S. Jolley, C.S. Rangel

A B

Fig. 4 T versus B cells: (A (left): Localization of T cells with sheep red blood cell receptor using Leu 5 anti· body. These cells are predominant in the PALS and MZ. Magnification 100 x. B (right): Localization of B cells using anti-kappa antibody on serial section of the same white pulp region . These cells are predominant in the GC, M, MZ and adjacent RP . Magnification 1OOx. Arrows indicate cells with natural peroxidase ac· tivity

talization in one particular white pulp site. ing antigen trapping and transfer to the ger­ ln Fig. 4A the antibody directed at mature minal center, there is clonal expansion of B T cells discloses prominence of T cells in the cells within the germinal center where abun­ periarteriolar lymphoid sheath and mantle dant mitoses are noted (2, 5- 7). zone with rare cells in the germinal center The findings in Figure 4 substantiate striking and red pulp. This particular monoclonal an­ compartmentalization of B and T cells in the tibody (Leu 5) detects T cells with the sheep white pulp of the spleen. T cells occur mainly red blood cell receptor (E rosette receptor). in the PALS and MZ but they are also admixed These cells correspond closely to those meas­ with B cells in the mantle region of the germi­ ured by standard cell suspension rosetting. ln nal center and within the substance of the Fig. 4B the antibody detects surface and cyto­ germinal center itself. The occasional T cell plasmic immunoglobulin-bearing cells. This noted in the germinal center corresponds with staining pattern confirms that surface immuno­ our findings and that of others in lymph nodes globulin-bearing cells are located primarily in (8). T cells, therefore, appear to play a funda­ the germinal center, mantle of the germinal mental role in B cell proliferation and germinal center, marginal zone and immediately adja­ center formation (2). The overlap of T and B cent red pulp with few, if any, in the PALS. cell zones in the white pulp of the spleen prob­ It is likely that these B cells home to the ably reflect the complex functional interaction mantle zone and the mantle of the germinal of B and T cells in this region. center from the vasculature (2, 5, 6). Follow-

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. Immunoarchitecture of the Human Spleen 77

A B

Fig. 5 Immunoglobulin-bearing cells : A (left) : Localization of kappa light chain-bearing cells using anti·kappa antibody. B (right) : Localization of lambda light chain-bearing cells using anti-lambda antibody. Note that kappa outnumber lambda bearing cells and some GC cells are immunoglobulin negative. Magnification 60 x

C Localization of immunoglobulin-bearing lymph nodes that the predominant germinal cells: center cell is surface immunoglobulin negative Figures SA and B demonstrate an abundance and immature B cell antigen positive (9). The of surface and/or cytoplasmic immunoglobulin­ majority of these germinal center cells repre­ bearing cells of both kappa and lambda light­ sent less mature B cells (9). Figure 5 illustrates chain types in the white pulp. In each case a that this phenomenon is "zonal" with most surface polyclonal pattern of both kappa and lambda immunoglobulin positive cells clustered at one light-chain expression is apparent. Kappa-bear­ end of the germinal center, probably reflecting ing cells, as illustrated, consistently outnum­ zonal maturation within the germinal center (9). ber lambda-bearing cells as previously described Abundant mitoses are noted in the germinal in reactive lymph nodes (9). The immunoglobu­ centers indicating frequent clonal expansion lin-bearing cells are located primarily in the of B cells. Throughout the B cell zone, large mantle of the germinal center, in the margi­ transformed immunoblasts are noted along nal zone and in the adjacent red pulp with a with differentiated plasma cells. Plasma cells few visible in the PALS. Note that the germi­ are densest at the edge of the germinal center nal center demonstrates mainly "intracellular" suggesting frequent B cell differentiation in deposition of immunoglobulin but with mini­ this zone. In one spleen, plasma cells are also mal cellular surface immunoglobulin apparent, prominent in the splenic red pulp. Thus, all corresponding with earlier observations in the known proliferative, transforming, and

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. 78 Th.M. Grogan, C.S. Jolley, C.S. Rangel

A B

Fig. 6 Subset T cells. A (left) : Localization ofT helper cells using Leu 3a antibody. B (right): Localization of T suppressor/cytotoxic cells using Leu 2a antibody. T helper cells outnumber T suppressor/cytotoxic cells in the PALS and MZ. Magnification 100 x differentiating stages of B cell ontogeny are the predominant functional T cell subset in represented in the reactive B cell zones of the splenic white pulp, a fmding observed in the splenic white pulp. all six cases studied. As illustrated T helper cells generally exceed T suppressor/cytotoxic D. Localization of subset T cells: cells by a substantial ratio estimated in excess As depicted in Figs. 6 and 7, monoclonal of 4 to 1 as we and others have observed in antibodies to T cell subsets demonstrate the reactive lymph nodes (8). distribution of subset cells in the white pulp. Fig. 6 demonstrates the occurrence of T hel­ The Leu 3a identifies T helper cells (Fig. 6A), per cells in germinal centers. Full B cell de­ and the Leu 2a identifies T suppressor/cyto­ velopment requires the presence ofT helper toxic cells (Fig. 6B). The assessment of sub­ cells (2), and this close proximity of B and T set T antigens in Figure 6 was performed on helper cells in germinal centers probably facil­ serial sections, allowing determination of itates their functional interaction. the distribution and number of subset T cells The distribution ofT suppressor/cytotoxic in a single white pulp location. cells is shown in 68. In the white pulp, these The staining pattern in Fig. 6A indicates the cells predominate in the PALS but are absent majority of cells in the PALS and MZ are T in the germinal center. As mentioned previous­ helper cells with only a scattering of these ly , there are considerably fewer T suppressor/ cells in secondary germinal centers and red cytotoxic cells than T helper cells in this re­ pulp. Compared to Fig. 68, T helper cells are gion. While T suppressor/cytotoxic cells are a

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. lmmunoarchitecture of the Human Spleen 79

A B

Fig. 7 T suppressor/cytotoxic cells are scattered throughout the red pulp. Antibody Leu 2a. Magnification 60 X, 100 X minority component in the white pulp, they cytic enzyme activity shown in Figs. 8A and B are the predominant T cell component in the demonstrates the prominence of monocytic/ red pulp (Fig. 7), where their distribution histiocytic cells in the red pulp. These cells corresponds closely to that of histiocytes, are visible both within cords and lining sinuses. monocytes and natural killer (NK) cells (Figs. While an occasional histiocyte is also seen in 8 and 9). germinal centers and marginal zones, the stain­ Previously, immune function in the red pulp ing pattern confirms that the predominant pha­ was linked primarily to phagocytic, pitting or gocytic activity of the spleen is in the red pulp. culling functions and in turn to monocytic/ Some of the monocytic/histiocytic enzyme po­ histiocytic activity (6). Our findings suggest sitive cells in the germinal center and marginal that T suppressor/cytotoxic cells may operate zone probably are antigen presenting cells. In in conjunction with histiocytes to effect the one case, special enzyme inhibition studies us­ cytotoxic and phagocytic function of the red ing sodium fluoride indicate that monocytes pulp. NK cells are also noted in close associa­ are particularly prominent in the marginal tion so that a triumvirate of natural killer, zone region while histiocytes are more promi­ T suppressor/cytotoxic and histiocytic cells nent in the cords and sinuses. The fmding of may collaborate in red pulp functions. the more immature monocytic elements in the marginal zone, which is rich in terminal arterio­ E. Localization of monocytes and histiocytes: les (Fig. 2 and 3), may signify that these cells are either recent arrivals from the bloodstream Tissue section assessment of monocytic/histio- or are pivotal in antigen trapping at this site.

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. 80 Th.M. Grogan, C.S. Jolley, C.S. Rangel

A B

Fig. 8 Monocytic and histiocytic cells are abundant in the red pulp. Magnification A (25 x), B (100 x) Butyrate esterase sta in

As previously mentioned, the more mature Historically, the spleen has been regarded as histiocytes predominate in the red pulp cords a central site of NK cell activity (1 0). Specific and exhibit a distribution similar to T sup­ antibodies to NK cells and T suppressor/cyto­ pressor/cytotoxic cells. However, they more toxic cells now distinguish between these over­ conspicuously line the sinuses while T sup­ lapping cytotoxic functional cells. In the pres­ pressor cells are more restricted to the cords. ent study these antibodies indicate a great pre­ dominance of T suppressor cells over true NK cells in the red pulp, suggesting that cytotoxic F. Localization of natural killer cells: and cytolytic functions of the spleen relate The monoclonal antibody directed at natural more toT suppressor/cytotoxic cells than to killer cells (NK cells) indicates they are found NK cells. Whether this is also true in pathologic in two splenic locations: 1) scattered about conditions is unknown. the red pulp and 2) conspicuously restricted to the germinal centers in the white pulp IV. Conclusion (Fig. 9). The latter pattern corresponds to the The tissue section method described here re­ NK restriction in germinal centers that we presents a new and exciting way of assessing have observed in reactive lymph nodes and the immunoreactivity of the spleen. Using snap . NK cells occur in substantially fewer frozen tissue sections, hybridoma-derived monc numbers than T suppressor/cytotoxic cells clonal antibodies and biotin-avidin methods of and histiocytes in the red pulp. Thus, while conjugation along with peroxidase labelling, NK cells may function in a triad with T sup­ the B and T cell zones of the spleen can be pressor/cytotoxic and histiocyte cells, they fully delineated. T cells predominate in the constitute a minority component in the nor­ periarteriolar lymphoid sheaths and B cells mal function of the spleen. in the germinal centers and mantle region.

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. lrnmunoarchitectUie of the Human Spleen 81

A B

Fig. 9 Natural Killer (NK) cells, A (left) are scattered in the red pulp and within germinal centers. Antibody Leu 7. Magnification 25x. B (right):Detail of NK cells in a germinal center. Magnification 250 x

There is nonetheless considerable overlap of the red and white pulp remain unexplained, these lymphocytes in the mantle zone as well the tissue section mode of analysis is likely as spillover of both into the adjacent red pulp. to prove useful in further delineating the fas­ Among lymphoid subsets, T helper cells strik­ cinating details of the immunoarchitecture ingly predominate overT suppressor/cytotoxic of the spleen in normal and diseased states. cells in the white pulp. Some T helper cells reside in germinal centers indicating a close References relationship between B clonal expansion and 1 Warnke, R., R. Levy: Detection ofT and B cell T helper cell activity. While T suppressor/cy­ antigens with hybridoma monoclonal antibodies: toxic cells are a minority T cell component a biotin-avidin-horseradish peroxidase method. in the white pulp, they are the predominant J. Histiochem. and Cytochem. 28 (1980) 771-776 2 Weissman, I.L. , R . Warnke, E.C. Butcher, R. T cell component in the red pulp where they Rouse, R. Levy: The lymphoid system, its nor­ are closely admixed with NK cells and mono­ mal architecture and the potential for understand­ cytic/histiocytic cells. This close proximity ing the system through the study of lymphopro­ suggests that they function as a triad in effect­ liferative diseases. Hum. Pathol. 9 (1978) 25 - 33 3 Wood, T., R. Warnke: Suppression of endogenous ing the culling and pitting functions of the avidin-binding activity in tissues and its relevance spleen. to biotin-avidin detection systems. J. Histiochem. Collectively, our findings demonstrate the Cytochem. 29 (1981) 1196- 1204 4 Yam, L.T., C.A. Li, W.H. Crosby: Cytochemical anatomic compartmentalization of lymphoid identification of monocytes and granulocytes. Am. and histiocytic cells in the spleen. Although J. Clin. Pathol. 55 (1971) 283 - 290 the complex overlapping patterns in both 5 Weiss, L.: The Spleen. ln the Blood Cells and

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY. 82 Th.M. Grogan, C.S. Jolley, C.S. Rangel

Hematopoietic Tissues. Chapter 15, pp. 545 - cell subsets in human lymph nodes. J. Exp. Med . 573. McGraw-Hill, New York, N.Y. 1977 153 (1981) 30- 41 6 Crosby, W.H. : Structure and Function of the 9 Stein, H., A. Bonk, G. Tolksdorf et al.: Immuno­ Spleen. In Hematology. Chapter 9, pp. 74- 81. histologic analysis of the organization of normal Eds.: Williams, W.J., E. Beutler, A.J. Ersleu, lymphoid tissue and non-Hodgkin's . R.W. Rundles, 2nd Ed., McGraw-Hill, New York, J. Histochem. Cytochem. 28 (1980) 746- 760 1977 10 Gupta, S., C. Tan : Subpopulations of human T 7 Langevoort, H.L.: The histophysiology of the an­ lymphocytes. XIV. Abnormality ofT-cell locomo­ tibody response. I. Histogenesis of the plasma tion and of distribution of subpopulations of T cell reaction in rabbit spleen. Lab. Invest. 12 and B lymphocytes in peripheral blood and spleen (1963) 106- 116 from children with untreated Hodgkin's disease. 8 Poppema, S., A.K. Bhan, E.L. Reinlzerz, R.T. Clin. Immunol. and Immunopathol. 15 (1980) McCluskey, S.F. Schlossman: Distribution of T 133 - 143

Tlz .M. Grogan, M.D., C.S. Jolley, M.T. , C.S. Rangel, H.T.L., Department of Pathology, University of Arizona, Arizona Health Sciences Center, Tucson, Arizona 85724

Permission granted for single print for individual use. Reproduction not permitted without permission of Journal LYMPHOLOGY.